System level modeling and closed loop control for a droplet-based micro-actuator

Micro electro-mechanical systems (MEMS) have drawn increasing interest due to the still ongoing miniaturization. Nevertheless, only a few closed loop control approaches have been developed for this class of systems. Therefore a combined data driven and classical modeling approach is proposed to enable the use of control techniques for such systems. Using system identification the suitability of this approach for a droplet-based actuator is demonstrated. In this specific case the analytical derivative of the incorporated neural network is harnessed to model a capacitance gradient. Finally a flatness-based closed loop control is deduced and tested in simulation to show the applicability of the proposed model within a closed loop control.